Metabolic Phenotyping from Whole-Blood Responses to a Standardized Exercise Test May Discriminate for Physiological, Performance, and Illness Outcomes: A Pilot Study in Highly-Trained Cross-Country Skiers

Abstract

BackgroundThis study used metabolic phenotyping to explore the responses of highly-trained cross-country skiers to a standardized exercise test, which was part of the athletes’ routine testing, and determine whether metabolic phenotyping could discriminate specific physiological, performance, and illness characteristics.MethodsTwenty-three highly-trained cross-country skiers (10 women and 13 men) participated in this study. Capillary whole-blood samples were collected before (at rest) and 2.5 min after (post-exercise) a roller-ski treadmill test consisting of 5–6 × 4-min submaximal stages followed by a self-paced time trial (~ 3 min) and analyzed using mass spectrometry. Performance level was defined by International Ski Federation distance and sprint rankings. Illness data were collected prospectively for 33 weeks using the Oslo Sports Trauma Research Center Questionnaire on Health Problems. Orthogonal partial least squares-discriminant analyses (OPLS-DA) followed by enrichment analyses were used to identify metabolic phenotypes of athlete groups with specific physiological, performance, and illness characteristics.ResultsBlood metabolite phenotypes were significantly different after the standardized exercise test compared to rest for metabolites involved in energy, purine, and nucleotide metabolism (all OPLS-DA p < 0.001). Acute changes in the metabolic phenotype from rest to post-exercise could discriminate athletes with: (1) higher vs. lower peak blood lactate concentrations; (2) superior vs. inferior performance levels in sprint skiing, and (3) ≥ 2 vs. ≤ 1 self-reported illness episodes in the 33-week study period (all p < 0.05). The most important metabolites contributing to the distinction of groups according to (1) post-exercise blood lactate concentrations, (2) sprint performance, and (3) illness frequency were: (1) inosine, hypoxanthine, and deoxycholic acid, (2) sorbitol, adenosine monophosphate, and 2-hydroxyleuroylcarnitine, and (3) glucose-6-phosphate, squalene, and deoxycholic acid, respectively.ConclusionMetabolic phenotyping discriminated between athlete groups with higher vs. lower post-exercise blood lactate concentrations, superior vs. inferior sprint skiing performance, and more vs. less self-reported illnesses. While the biological relevance of the identified biomarkers requires validation in future research, metabolic phenotyping shows promise as a tool for routine monitoring of highly-trained endurance athletes.